NASA Technical Memorandum 87090

Thermal Desorption Study of PhysicalForces at the PTFE Surface

(NASA-TM-87090) 1HEEMAL DESCFEUCN STUDY OF N86-25284PHYSICAL FCRCIS AT TEE FTff SUBFACE JNASA)15 p HC A02/M* AO1 CSCL 20C

DnclasG3/77 43479

D.R. Wheeler and S.V. PepperLewis Research CenterCleveland, Ohio

Prepared for theInternational Conference on Surface and Colloid Chemistrysponsored by the American Chemical SocietyPotsdam, New York, June 24-28, 1985

NASA

https://ntrs.nasa.gov/search.jsp?R=19860015813 2020-02-22T11:58:32+00:00Z

THERMAL DESORPTION STUDY OF PHYSICAL FORCES AT THE PTFE SURFACE

D.R. Wheeler and S.V. PepperNational Aeronautics and Space Administration

Lewis Research CenterCleveland, Ohio 44135

SUMMARY

Thermal desorptlon spectroscopy (TDS) of the polytetraf luoroethylene(PTFE) surface was successfully employed to study the possible role of physi-cal forces 1n the enhancement of metal-PTFE adhesion by radiation. The thermaldesorptlon spectra were analyzed without assumptions to yield the activationenergy for desorptlon over a range of xenon coverage from less than 0.1 mono-layer to more than 100 monolayers. For multilayer coverage, the desorptlon 1szero-order with an activation energy equal to the sublimation energy of xenon.For submonolayer coverages, the order for desorptlon from the unlrradlatedPTFE surface 1s 0.73 and the activation energy for desorptlon 1s between 3.32and 3.36 kcal/mol; less than the xenon sublimation energy. The effect ofIrradiation 1s to Increase the activation energy for desorptlon to as high as4 kcal/mol at low coverage.

INTRODUCTION

The adhesion between polytetrafluoroethylene (PTFE) and metals Is ofpractical Importance. Good dielectric and thermal properties make PTFE useful1n electronic applications where 1t 1s desirable to Increase the adhesion ofmetal films to the PTFE. In many mechanical applications a strong bond 1srequired between a metal substrate and a PTFE film, and 1n trlbologlcal appli-cations, 1t 1s necessary that the transfer film of PTFE formed during slidingadhere well to the metal counterface. In all these cases, 1t would be desira-ble to Increase the normally low adhesion between metal and PTFE.

It has been found that metal films adhere better to Irradiated PTFE thanto virgin PTFE, whether the radiation 1s Ions (ref. 1), electrons (ref. 2), orx-rays (ref. 3). Understanding this Improvement could help 1n understanding :the bond between metals and unlrradlated PTFE. The Improved adhesion onIrradiated PTFE can be due to one or more of three types of effect (ref. 4);topographic changes (e.g., mechanical Interlocking), chemical Interactions, orphysical (dispersion) forces. Mechanical forces are unlikely 1n the presentcase, because no transfer of the PTFE to the metal 1s observed when the bondfalls (ref. 1). Furthermore, the enhanced adhesion is observed upon irradi-ation with either x-rays or ions which produce substantially different topo-graphic changes 1n the surface of the PTFE (ref. 3). There 1s some x-rayphotoelectron spectroscoplc (XPS) evidence for chemical interaction betweennickel films and irradiated PTFE (ref. 3). However the data could not beInterpreted unambiguously, and since the PTFE is damaged by the x-rays usedfor analysis, the technique 1s open to question. Physical forces are alwayspresent between materials. While much weaker per atom than chemical forces,they act over the entire contact area and can thus be an appreciable part ofthe total bond between macroscopic surfaces. The question is whether the phys-ical forces are Increased by Irradiation and whether they can account for theincreased adhesion. To date, there is no evidence on either of these points.

Inert gas adsorption 1s commonly used as a probe of physical Interactionson surfaces. One of the usual techniques 1s thermal desorptlon spectroscopy(IDS) (ref. 5). To the best of our knowledge, 1t has not been used on polymersurfaces. The purpose of the present study was twofold; first, to develop theTDS technique for xenon desorptlon from planar polymer surfaces and second touse the technique to determine the change 1n physical forces at the PTFE sur-face upon Irradiation with electrons.

EXPERIMENT

• Thermal desorptlon experiments were performed 1n an ultrahlgh vacuumchamber fitted as shown schematically 1n figure 1. The chamber was pumpedwith a 150 L/s turbopump and a titanium sublimation pump. A nude 1on gauge,out of the line of sight of the specimen, was used to monitor the backgroundpressure 1n the chamber and also to measure the pressure rise during the TDSexperiments. The pressure 1n the chamber, before beginning a TOS experiment,was less than BxlQ-11 torr. A cryostat regulated the sample temperature.Xenon was directed onto the specimen through a mlcrocaplllary array from acalibrated ballast volume. The chamber also Incorporated an electron gun forIrradiation of the sample and apparatus (not shown 1n figure 1.) for x-rayphotoelectron spectroscopy (XPS) of the specimen surface.

A TOS experiment consisted of exposing the specimen at 30 K to a givendose of xenon. The temperature of the specimen was then Increased linearly to120 K at a rate of 0.1 K/s, and both the specimen temperature and the pressure1n the chamber were recorded. Under the rapid pumping speed conditions ofthis experiment, the pressure rise 1s proportional to the desorptlon rate ofxenon from the PTFE surface (ref. 6). The resulting curve of pressure versustemperature 1s the thermal desorptlon spectrum. The pressure, temperaturepairs were digitized and stored In a microcomputer. Some details of the spec-imen preparation, the dosing procedure and the cryostat are presented here.

Specimen Preparation

The PTFE surfaces used for TOS were prepared by spinning a commercialdispersion of PTFE onto copper caps which could be screwed onto the cryostatused to regulate the sample temperature. Before use, the coated caps wereannealed 1n vacuum to between 410 and 415 °C for 15 m1n to drive off the car-rier and sinter the PTFE film. Scanning electron microscopy of the specimensshowed the filamentary structure typical of specimens prepared 1n this way(ref. 7). They undoubtedly had a surface area greater than the geometric area.

Irradiated PTFE samples were prepared 1n the same way. They were then,exposed to a 1 kV electron beam. The beam was rastered over the surface ofthe specimen. The effective current density was 2.2 yA/cm2, and the samplewas Irradiated for 180 m1n. Irradiation was performed 1n the vacuum system 1nwhich the TOS experiments were performed, and the sample was not exposed toair between Irradiation and TDS analysis.

Samples prepared for TDS were examined with XPS before and after Irradi-ation. Because the x-ray Irradiation used to obtain the XPS spectra causesdamage to the PTFE (ref. 8), samples that were used for TDS experiments werenot analyzed by XPS until after the TDS experiments were complete. The XPS

spectrum of the unlrradlated PTFE specimens was characteristic of clean,undamaged PTFE. No Impurities could be detected, and there was no trace ofcopper or oxygen lines from the copper substrate. The change upon Irradiationwas the same as had been observed previously (ref. 8).

Ooser

The apparatus used to dose the specimen with xenon consisted of a micro-capillary array connected through a bakable valve to an 1on pumped ballastvolume. Pressure 1n the ballast volume was monitored with an ton gauge. Theamount of xenon to which the specimen was exposed was calculated from the drop1n pressure 1n the known ballast volume during dosing. From the area of thespecimen exposed to xenon and the diameter of the xenon atom, the number ofmonolayers which would result from a particular amount of xenon on a geomet-rically smooth surface was calculated. In the data to follow, that number ofmonolayers 1s referred to as the dose.

The actual coverage of xenon on the PTFE after dosing depended on thetrue surface area of the specimen and the sticking coefficient of the xenon aswell as on the dose. It has already been noted that the true surface area wasgreater than the apparent area. Furthermore, during dosing, there was a slightpressure rise 1n the main chamber which Indicated that the sticking coefficientwas less than one. However, the pressure rise and the time 1t lasted were bothmore than an order of magnitude less than the pressure Increase and time of aTDS. Thus, the sticking coefficient must be greater than 0.99. In any case,the actual xenon coverage was proportional to the area under a TDS curve. Inall the data below, the coverage of xenon 1s calculated from the area under theTDS spectrum. The unit 1s torr-sec. As a point of reference, the coverageproduced by a xenon dose of one monolayer was 6.1xlO~^ torr-sec. That cov-erage 1s Indicated as 1 ml 1n the data below.

In principle, the thickness of the xenon layer on the PTFE could be meas-ured by the attenuation of the XPS lines from the PTFE or by angle resolvedXPS. Both methods were tried, but 1t was found that the radiation used forXPS was sufficient to cause progressive damage to the PTFE during the analysis.This produced variable xenon adsorption. Furthermore, the radiation damagedPTFE was not stable during the rather long times required for XPS analysis.Finally, the Inelastic mean free path of low energy electrons 1n xenon, which1s required for the analysis, 1s not well known. As a result, the true mono-layer coverage of xenon could not be found precisely. It was found that adose of 1 ml produced a coverage of between 0.2 and 0.5 ml as determined byXPS.

Cryostat

The sample was mounted on a continuous flow liquid He refrigerator. Therefrigerator Incorporated a heater for temperature control and a thermocouplefor temperature measurement. Because the thermocouple was separated from thesample surface, the actual sample temperature could differ by several tenthsof a degree from the Indicated temperature. Furthermore, the discrepancy var-ied throughout a day of operation. The temperature of the peak 1n the TDSfrom a standard dose of xenon was used to correct for this temperature drift.

Thus, temperatures reported here are consistent to within 0.2 K, but the abso-lute temperature error could be as large as 0.5 K.

It was found that readsorptlon of the xenon from the specimen onto othercold areas of the cryostat produced serious artifacts 1n the IDS. This was aparticular problem at the low ramp rate of 0.1 K/sec used here. Readsorptlonwas controlled by adding a room temperature shield to the cryostat as shown 1nfigure 1. Clearance between the shield and the cold specimen 1s minimal, andthe amount of desorbed xenon reaching other parts of the cryostat 1s notdetectable in the spectrum. Furthermore, to assure that the cryostat andspecimen are cleared of xenon, the specimen temperature Is raised to 220 Kbetween runs.

RESULTS

Thermal desorptlon spectra were obtained for a wide range of xenon doseson virgin and Irradiated PTFE. We will consider the general features of thespectra first and then turn to an analysis of the thermodynamlc realtlonshlpbetween the pressure, temperature and xenon coverage represented by each pointon a spectrum.

General Features

The TDS curves fell naturally Into two families; those for multilayerInitial coverage and those for submonolayer Initial coverage. Each familywill be considered separately.

Multilayer coverage. The general shape of the TOS for Initial coveragegreater than 1.3xlO~* torr-sec (about 21 ml) 1s Illustrated by the spectrumof figure 2(a). There was a characteristic rapid decrease 1n pressure at hightemperature: Figure 2(b) shows the low temperature (high coverage) region oftwo such spectra. The pressures and therefore the desorptlon rates 1n thistemperature range were the same for a wide range of coverage. The variation1n temperature of the TDS peak maximum 1s shown 1n figure 3. In the region ofmultilayer coverage, the peak temperature Increased with the coverage. All ofthese observations are consistent with zero-order desorptlon kinetics (ref. 9).In addition, figure 3 shows that the TDS peak temperature was Independent ofsubstrate Irradiation, 1n the multilayer coverage region.

Low coverage. Spectra obtained at submonolayer coverages are shown 1nfigures 4(a) and (b). They did not exhibit shapes characteristic of any simpledesorptlon model. In particular, they did not appear to be produced by first-order desorptlon. This 1s confirmed by the variation of the temperature ofthe peak maximum shown 1n figure 3. At submonolayer coverage, the peak tem-perature decreased with Increasing coverage. In the case of simple, first-order desorptlon, the temperature of the peak would remain constant. Thevarying peak temperature could be attributed either to a distribution of bind-ing energy sites on the surface or to fractional order desorptlon produced bylateral Interactions between xenon atoms (ref. 9).

Comparison of the high temperature sides of the two peaks 1n figure 4(a)suggests that Irradiation produced some Increase 1n the number of high energybinding sites. The effect of radiation on the substrate 1s even more apparent

1n figure 4(b). The TDS peak on the Irradiated specimen was almost 10 K higherthan the peak on the unlrradlated specimen. The effect of radiation on thesubstrate 1s also evident 1n the submonolayer coverage region of figure 3.Below 2xlO~7 torr-sec, the maximum pressure always occurred at higher tem-peratures for the Irradiated PTFE than for the unlrradlated PTFE.

Thermodynamlc Analysis

Because we wish to determine the strength of the xenon-PTFE Interaction,we are particularly Interested 1n the low coverage data. Had the desorptlonbeen first-order, a simple Redhead analysis would yield the desorptlon energy(ref. 6). However, 1n the present case we resort to analysis based explicitlyon the Arhenlus equation (ref. 5):

r = «Cn exp (-E/RT), (1)

where r 1s the desorptlon rate, v 1s the "frequency factor," C the cover-age, n the order of desorptlon, E the activation energy for desorptlon, Rthe gas constant and T the temperature. Most analyses of TDS assume thatv 1s Independent of coverage, and further, that 1t has a value of 10^ $-1.We do not make these assumptions but proceed as follows.

The pressure at each point 1n a TDS 1s proportional to the desorptlonrate of xenon from the surface. This rate depends on both the temperature andthe coverage at that temperature. The temperature can be determined directly.The coverage 1s proportional to the Integral under the TDS curve from the tem-perature of Interest to the highest temperature. On each TDS, temperature,pressure and coverage values were obtained at half degree Intervals up to thepeak of the curve. The pressure and coverage at one temperature were measuredon TOS curves for a variety of Initial doses. These were combined to producea plot of pressure versus coverage at that temperature.

A typical plot for virgin PTFE 1s shown 1n figure 5. The logarithmiccoverage scale allows presentation of a wide range of coverage but has noother significance. At very low coverage, the desorptlon rate must, ofcourse, approach zero. As the coverage Increases, so does the desorptlonrate, but the Increase 1s not linear as 1t would be for first-order desorp-tlon. At a coverage of 9xlO~6 torr-sec (15 ml), the desorptlon rate dropsand then remains constant for higher coverage. Coverage Independent desorp-tlon rate 1s the signature of zero-order desorptlon kinetics.

Curves of pressure versus coverage were constructed every half-degree fortemperatures between 50 and 60 K. Below 50 K, the pressure rise was not largeenough to be measured reliably, while there were too few data points above60 K. Reading the pressure at a particular coverage from each curve 1n thisseries yields the pressure as a function of temperature for that coverage. AnArhenlus plot of the natural log of the pressure versus the Inverse of thetemperature can then be made. Figure 6 1s a typical Arhenlus plot. The line-arity of these plots 1s confirmation that equation (1) holds with a singleenergy 1n the range of temperature and coverage analyzed. The slope of thebest line through the data 1s -E/R.

Arhenlus plots were constructed for a range of coverages, and the desorp-tlon energies determined. The result 1s shown 1n figure 7. For reference,

the sublimation of xenon 1s also shown (ref. 10). As can be seen, the desorp-tlon energy at high coverage was 1n reasonable agreement with the sublimationenergy. As the coverage decreased, the desorptlon energy decreased until 1treached a value of 3.34*0.03 kcal/mol below 1 monolayer coverage.

The same analysis was performed on the IDS from the Irradiated PTFE sur-face. Because the surface was not stable, fewer TOS curves could be acquired.However, some desorptlon energies could be extracted 1n the coverage rangefrom 10~7 to 10~5 torr-sec. These are also shown 1n figure 7. It can beseen that the energies were the same as those for virgin PTFE above 1 mono-layer. At low coverage, however, the desorptlon energy from Irradiated PTFEwas larger than that from virgin PTFE and exceeded even the sublimation energyof xenon.

A plot of the Intercepts of the Arhenlus plots versus the natural-log ofthe coverage 1s shown 1n figure 8. According to equation (1), the slope ofthis line 1s the order n of desorptlon which 1s 0.73 1n this case. Thelinearity of the plot confirms that the order and the preexponentlal, v, areconstants. There were Insufficient data for a similar analysis on theIrradiated surface.

DISCUSSION

To the best of our knowledge, this 1s the first report of the applicationof thermal desorptlon spectroscopy to a polymer surface. Therefore, the firstpart of the discussion will address the experimental technique, Itself. Next,conclusions will be drawn on the nature of the adsorption of xenon on the vir-gin PTFE surface. Finally, the effect of Irradiation of PTFE on xenon adsorp-tion and Its significance for adhesion will be discussed.

Experimental Technique

The experiment reported here differs in ramp speed from most reportedflash desorptlon experiments (refs. 5, 6, and 9). The low ramp speed wasrequired to assure temperature uniformity 1n the cryostat used. The uniformity1s demonstrated by the sharpness of the TDS curves. The sharp drop on thehigh temperature side of the multilayer peak 1n figure 2(a) demonstrates theresponse of which the system 1s capable. It also shows that the specimen tem-perature was uniform.

Because the ramp speed 1n this experiment was so low, the possibility ofreadsorptlon on the sample during the ramp must be considered (ref. 5). Infact, 1t was found that the Arhenlus plots displayed a slight departure fromlinearity at pressures above 5xlO~8 torr. This was attributed to effects ofreadsorptlon at high desorptlon rate, and pressures above this value were,therefore, excluded from the analysis. The linearity of the Arhenlus plots 1sone evidence for the validity of the data. The fact that the desorptlonenergy at high coverage was nearly equal to the sublimation energy of xenon 1sadditional evidence.

The nondestructive nature and surface sensitivity of the TOS give 1tunique capabilities for the analysis of surfaces such as that of PTFE. It wasfound that the TDS spectrum was sensitive to radiation-produced changes 1n the

surface that could not be detected by XPS. Indeed, the unstable nature of theIrradiated surface was only evident from the variability of the xenon desorp-tlon spectra.

Adsorpt1on/0esorpt1on on Virgin PTFE

For multilayer coverage, the IDS curves are characterized by peak-temperatures that Increase with coverage, Identical low temperature behaviorfor all coverages and sharp high temperature edges. These are all features ofzero-order desorptlon. This 1s confirmed by plots of pressure versus coveragewhich show that the pressure (and hence the desorptlon rate) were Independentof coverage above 15 monolayers. Zero-order desorptlon kinetics can be pro-duced by a variety of processes,11 the most obvious being sublimation frombulk xenon. In the present case, the agreement between the desorptlon energyat high coverage and the sublimation energy of xenon confirms the obviousInterpretation.

For submonolayer coverage, the energy of desorptlon was less than thesublimation energy of xenon. Clearly, the xenon-PTFE Interaction energy wasless than the xenon-xenon Interaction energy. The low energy of Interactionwith the substrate and the fractional order of the desorptlon make 1t likelythat lateral Interactions were affecting the desorptlon energy.9,11 it seemslikely that xenon Islands were present on the PTFE surface. In that case, thedesorptlon energy can only be an upper limit to the single-atom, xenon-PTFEInteraction energy.

The plot of pressure versus coverage 1n figure 5 clearly shows an abruptchange 1n desorptlon rate 1n the Intermediate range of coverage between oneand several monolayers. Without further evidence, 1t 1s not possible toInterpret this transition 1n detail, but 1t 1s certainly related to the changefrom desorptlon Influenced by the substrate to sublimation of bulk xenon.

Effect of Irradiation

At coverage less than 1 monolayer, the desorptlon energy of xenon fromIrradiated PTFE was 3.85 to 4.0 kcal/mol, whereas 1t was only 3.32 to 3.36kcal/mol from virgin PTFE 1n the same coverage range. Furthermore, the TDSpeaks were 10 K higher for the Irradiated substrate. It seems clear that theInteraction between xenon and Irradiated PTFE was stronger than between xenonand virgin PTFE. It 1s unlikely that chemical bonding was Involved betweenxenon and PTFE, so the Increased Interaction 1s attributed to Increased dis-persion forces.

What changes can Irradiation produce 1n PTFE that would enhance thesedispersion forcesy The changes 1n the C(ls) XPS spectrum of Irradiated PTFEwere previously Interpreted as evidence for crossllnking 1n the surface regionof the polymer.^ The density of the surface region could be Increased bycrossllnking. The Increased density would, 1n turn, lead to Increased disper-sion forces.12 However, the Instability of the Irradiated PTFE surface whenprobed by TDS suggests that the change 1s not entirely structural. Irradi-ation 1s known to produce trapped, long-lived radicals 1n PTFE.13 It seemsprobable that the optical polar1zab1!1ty of these radicals would be different

from that of PTFE. The presence of radicals then could well affect thedielectric properties and hence the dispersion forces at the PTFE surface.

Whatever the cause of the Increased dispersion forces, they can certainlycontribute to the adhesion of thin metallic films to PTFE. Since there alsoseem to be chemical differences 1n the effect of Irradiation on the adhesionof different metals.2 any explanation of the effect of radiation on adhesionmust take Into account both chemical and physical forces.

CONCLUSION

Thermal desorptlon spectroscopy of xenon from Irradiated and virgin PTFEhas been successfully performed. Analysis of the TDS curves and the desorp-tlon kinetics 1n the 50 to 60 K range showed the following

1. Oesorptlon from multilayers of Xe on either surface proceeded by zero-order kinetics with an activation energy equal to the Xe sublimation energy.

2. At a coverage below 3 to 7 true monolayers the desorptlon behaviorchanged abruptly from the multilayer case and was different on Irradiated andunlrradlated PTFE substrates.

3. For submonolayer coverages on Irradiated PTFE the activation energywas 3.85 to 4.0 kcal/mole, which was the highest observed.

4. For submonolayer coverages on unlrradlated PTFE the desorptlon energywas 3.34̂ 0.03 kcal/mole (less than the sublimation energy of Xe), and thedesorptlon was fractional order, Indicating significant lateral Interactions.

It can be concluded that the Xe-Xe Interaction was stronger than theXe-PTFE Interaction on unlrradlated PTFE. Irradiating PTFE Increased thedispersion forces between Xe and PTFE.

REFERENCES

1. J.E.E. Baglln, G.J. Clark, and J. Bottlger, 1n "Thin Films and InterfacesII," J.E.E. Baglln, et al., eds., pp. 179-188, North Holland, New York,1984.

2. Y. Yamada, D.R. Wheeler, and D.H. Buckley, NASA TP-2360 (1984).

3. D.R. Wheeler and S.V. Pepper, 0. Vac. Sd. Technol., 20, 442-443 (1982).

4. K.L. MHtal, Polymer Eng. Sc1., U, 467-473, (1977).

5. D.A. King, Surf. Sd., 47, 384-402 (1975).

6. P.A. Redhead, Vacuum, 1_2, 203-211 (1962).

7. O.C. Bassett and R. Davltt, Polymer, ]5_, 721-728 (1974).

8. D.R. Wheeler and S.V. Pepper, J. Vac. Sc1. Technol., 20, 226-232 (1982).

. 8

9. R.G. Jones and O.L. Perry, Surf. Sc1., 82, 540-548 (1979).

10. C.W. Lemlng and G.L. Pollack, Phys. Rev., B,2, 3323-3330 (1970).

11. R. Opllla and R. Gomer, Surf. Sc1., 112. 1-22 (1981).

12. K. Hara and H. Schonhorn, J. Adhesion, 2, 100-105 (1970).

13. R.E. Florin and J. Wall, J. Res. Nat. Bur. Stand. Sect A, 65A, 375-387(1961).

TITANIUMSUBLIMATION,/PUMP

ROOM TEMPERATURE SHIELD

TURBO-MOLECULAR PUMP

Figure 1. - Schematic representation of the thermal desorptionspectroscopy apparatus.

inincc:a.

30 r—

25

20

15

10

5

ccin

ono.

50

"4(a) Initial coverage of 1. 3xlO" torr-sec (21 monolayers).

21 ML

50 52 54 56TEMPERATURE, K

58

(b) Initial coverages of 4xlO"5 and 1.3xlO"4 torr-sec(6.5 and 21 monolayers, respectively).

Figure 2. - Thermal desorption spectra of xenon on PTFE.

-

10r6

.LU

§O

10'r7

10'

10,-9

— 1ML

OO

D

I I

OL

O

O UN1RRADIATEDD IRRADIATED

D

D

56 58 60 62 64 66 68 70IDS PEAK TEMPERATURE, K

D

I I I72 74

Figure 3.- Dependence of TDS peak maximum on initial coverage.

13 r~

12

11

UNIRRADIATED

(a) 3.3xlO~7 torr-sec initial coverage.

/-IRRADIATEDUNIRRADIATED A

i.ioCOUJ

£ 9

8

—

.

— ..-.-•"' ..•••' "'-.••.•.•;•..• "''''. 4'. .••••"""!' ''•'•'•'•••••»1 1 1

40 50 60 70 80 90TEMPERATURE, K

(b) 4.2xlO'9 torr-sec initial coverage.

Figure 4.- Thermal desorption spectra for irradiated andunirradiated PTFE substrates.

oo

</•>10

6

5

4

3

2

-8.5

-£>

IQD o

o° oo o o

r$>& I I I

-7.5 -6.5 -5.5 -45LOG OF XENON COVERAGE

-3.5

Figure 5. -Xenon pressure as a function of coverage at 52K,unirradiated PTFE surface.

-17

-18CO

ffl -19

fe -20

3 -21_i| -22i—H: _23

-2".0165 .0170 .0175 .0180 .0185 .0190 .0195 .0200

INVERSE TEMPERATURE, K"1

Figure 6. - Arhenius plot: Natural log of the pressure vsinverse temperature for xenon coverage of 1x10-5 torr-sec on unirradiated PTFE.

AC

TIV

ATI

ON

EN

ERG

Y FO

RD

ES

OR

PTI

ON

, kc

al/m

ole

4.0

3.8

3.6

3.4

3.2

1 n

O UNIRRADIATED PTFED IRRADIATED PTFE

— D

On— ^ XENON SUBLIMATION ENERGY .^

"~ C3DOO OOOOCD

— 1ML

1 1

O Q °°

8

110" 10',-7 10',-6 10' 10',-4

XENON COVERAGE, torr-sec

Figure 7. - Activation energy for desorption as a functionof xenon coverage in the temperature range of 50 K to60 K.

12

Q.

l/l11

< 10

£LLJO

25 9

I I-17.5 -16.5 -15.5 -145

NATURAL LOG OF COVERAGE-13.5

Figure 8. - Intercept of arhenius plots vs natural log of thexenon coverage unirradiated PTFE.

1. Report No.

NASA TM-870902. Government Accession No. 3. Recipient's Catalog No.

4. Title and Subtitle 5. Report Date

Thermal Desorptlon Study of Physical Forces at thePTFE Surface 6. Performing Organization Code

506-53-127. Author(s)

D.R. Wheeler and S.V. Pepper

8. Performing Organization Report No.

E-2680

10. Work Unit No.

9. Performing Organization Name and Address

National Aeronautics and Space AdministrationLewis Research CenterCleveland, Ohio 44135

11. Contract or Grant No.

12. Sponsoring Agency Name and Address

National Aeronautics and Space AdministrationWashington, D.C. 20546

13. Type of Report and Period Covered

Technical Memorandum

14. Sponsoring Agency Code

15. Supplementary Notes

Prepared for the International Conference on Surface and Colloid Chemistry,sponsored by the American Chemical Society, Potsdam, New York, June 24-28, 1985.

16. Abstract

Thermal desorptlon spectroscopy (TDS) of the polytetrafluoroethylene (PTF£) sur-face was successfully employed to study the possible role of physical forces 1nthe enhancement of metal-PTFE adhesion by radiation. The thermal desorptlonspectra were analyzed without assumptions to yield the activation energy fordesorptlon over a range of xenon coverage from less than 0.1 monolayer to morethan 100 monolayers. For multilayer coverage, the desorptlon 1s zero-order withan activation energy equal to the sublimation energy of xenon. For submonolayercoverages, the order for desorptlon from the unlrradlated PTFE surface 1s 0.73and the activation energy for desorptlon 1s between 3.32 and 3.36 kcal/mol: lessthan the xenon sublimation energy. The effect of Irradiation 1s to Increase theact1v1at1on energy for desorptlon to as high as 4 kcal/mol at low coverage.

17. Key Words (Suggested by Authorfs))

AdhesionRadiationPTFEThermal desorptlon spectroscopy

19. Security Classif. (of this report)

Unclassif ied

18. Distribution Statement

Unclassified -STAR Category

20. Security Classif. (of this page)

Unclassified

- unlimited77

21. No. of pages 22. Price*

*For sale by the National Technical Information Service, Springfield, Virginia 22161

National Aeronautics andSpace Administration

Lewis Research CenterCleveland. Ohio 44135

Official BusinessPenalty tor Private Use $300

SECOND CLASS MAIL

ADDRESS CORRECTION REQUESTED

Postage and Fees PaidNational Aeronautics andSpace AdministrationNASA-451

NASA